MASID-v3

MASID-v3 is a fine-tuned version of Qwen2.5-7B trained specifically for Filipino recipe generation, with a focus on main dish preparation.

This model was trained on the Filipino Recipes 2K V2 dataset, a curated collection of ~2,000 authentic Filipino recipes.
Unlike earlier variants that explored multi-stage fine-tuning, MASID-v3 was trained directly from Qwen2.5-7B using this dataset to specialize the model toward Filipino culinary knowledge.

The goal of MASID-v3 is to generate structured and culturally accurate Filipino main dish recipes, covering a wide range of traditional cooking methods and ingredient combinations.

---

Model Details

• Base Model: Qwen2.5-7B
• Dataset: Filipino Recipes 2K V2 (~2,000 samples)
• Training Objective: Recipe text generation (Filipino cuisine, main dishes)
• Method: Direct fine-tuning from Qwen2.5-7B

---

Intended Use

• Assisting in recipe writing
• Exploring Filipino food culture
• Generating cooking instructions in natural language

---

Limitations

• The model was trained on a relatively small dataset (~2k samples).
• May sometimes produce hallucinated ingredients or inaccurate cooking steps.
• Not suitable for use as a nutritional or food safety reference.
• Best used for research, education, and creative applications.

---

Evaluation

Dataset	Split	BLEU-4	METEOR	ROUGE-L (F1)
joackimagno/FILIPINO_RECIPES_2K_V2	test	0.07	0.35	0.32

---

---

This Qwen2 model was trained 2× faster with Unsloth and Hugging Face’s TRL library.

Example Usage

from typing import List
import torch
from transformers import AutoModelForCausalLM, AutoTokenizer, GenerationConfig

# Load model and tokenizer
model_name = "joackimagno/MASID-v3"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForCausalLM.from_pretrained(
    model_name,
    torch_dtype=torch.float16,
    device_map="auto",
)

# ==============================================================
# Alpaca-style prompt
# ==============================================================

SYSTEM_INSTRUCTION = (
    "You are a Filipino chef. Generate Filipino MAIN DISH recipes.\n"
    "Follow these output rules:\n"
    "1) Use standard stovetop or oven methods.\n"
    "2) Keep steps concise and logically ordered.\n"
    "3) Output FORMAT and ORDER must be exactly:\n"
    "   Recipe name, Prep time, Cook time, Total time, Servings,\n"
    "   Full Ingredients (numbered list), Instructions (numbered list)"
)

ALPACA_TEMPLATE = (
    "Below is an instruction that describes a task, paired with an input that "
    "provides further context. Write a response that appropriately completes the request.\n\n"
    "### Instruction:\n{}\n\n### Input:\n{}\n\n### Response:\n{}"
)

def make_model_input_from_ing(ing_names: List[str]) -> str:
    return (
        "Ingredients to use: " + ", ".join(ing_names) + ".\n"
        "Task: create a Filipino main dish recipe using these ingredients. "
        "Keep steps concise, clear, and coherent."
    )

# Example input
ing_names = ["Beef", "Potato", "Sili", "Carrot", "Sayote"]

alpaca_prompt = ALPACA_TEMPLATE.format(
    SYSTEM_INSTRUCTION,
    make_model_input_from_ing(ing_names),
    ""  # leave response empty for model to generate
)

# ==============================================================
# Run inference
# ==============================================================

inputs = tokenizer(alpaca_prompt, return_tensors="pt").to(model.device)

gen_config = GenerationConfig(
    max_new_tokens=512,
    temperature=0.7,
    top_p=0.9,
    do_sample=True,
)

outputs = model.generate(**inputs, generation_config=gen_config)

generated = tokenizer.decode(
    outputs[0][inputs["input_ids"].shape[1]:],
    skip_special_tokens=True
)

print(generated.strip())